scapesClassification is an R-package for the classification of seascapes, landscapes and other geo-spaces based on user-defined conditions. It allows users to translate task-oriented views of spaces and of geographic objects into computer representations.
Geo-spaces are classified using sets of user-defined conditions. These sets of conditions, presented in the form of conditional statements, can be applied simultaneously or sequentially. The second alternative is generally preferred. In fact, the rationale that guided the design of scapesClassification is better implemented considering multi-step classification processes:
Geographic objects and segments of space can be identified by a unique class (or set of classes);
Generally, it is easier to identify attributes that are distinctive of a portion of a class rather then identify its full range of attributes at once;
Thus, distinctive attributes can be used to map an initial set of locations to a class. These locations are hereafter referred to as anchor locations;
Then, it becomes possible to map a location to a class not only considering its intrinsic attributes, but also its spatial relationships with anchor locations;
In particular, class contiguity and class continuity can be considered;
Class contiguity. Qualitative and/or quantitative conditions define the membership of a location to a class only if they occur in positions considered adjacent to a specific class;
Class continuity. When two adjacent locations are assigned to the same class using contiguity conditions, these conditions can be re-applied at positions neighboring the newly classified locations;
Distinct classes can be identified repeating the above process and considering the relationships expected to exist among classified and unclassified locations.
The following is an example of how to use scapesClassification to classify marine seafloor structures, hereafter referred to as geomorphic management units (GMUs). Our study area is located in the ‘Triangle’ of the Azores (NE Atlantic), around the islands of Faial, Pico and São Jorge. GMUs will be identified using a RasterStack of six layers that includes bathymetry, bathymetric derivatives (Walbridge et al., 2018) and the geomorphons of the area (Jasiewicz and Stepinski, 2013):
We will start by loading the required libraries and data into the workspace:
# LOAD LIBRARIES
library(raster)
library(scapesClassification)
# LOAD DATA
grd <- list.files(system.file("extdata", package = "scapesClassification"), full.names = T)
grd <- grd[grepl("\\.grd", grd)]
grd <- grd[!grepl("hillshade", grd)]
rstack <- stack(grd)
We can plot the raster stack as an interactive map using the R packages mapview and leaflet:
FIGURE 1 - Raster layers used for the classification of Geomorphic Management Units (GMUs). The geomorphon classes correspond to (1) flat, (2) summit, (3) ridge, (4) shoulder, (5) spur, (6) slope, (7) hollow, (8) footslope, (9) valley and (10) depression.
OKEANOS-Azores University, Azores Deep Sea Research Group (@AzoresDeepSea), 9901‐862 Horta, Portugal. Email: gh.taranto@gmail.com↩︎